Pleiotropic effect of sigE over-expression on cell morphology, photosynthesis and hydrogen production in Synechocystis sp. PCC 6803

Authors


Summary

Over-expression of sigE, a gene encoding an RNA polymerase sigma factor in the unicellular cyanobacterium Synechocystis sp. PCC 6803, is known to activate sugar catabolism and bioplastic production. In this study, we investigated the effects of sigE over-expression on cell morphology, photosynthesis and hydrogen production in this cyanobacterium. Transmission electron and scanning probe microscopic analyses revealed that sigE over-expression increased the cell size, possibly as a result of aberrant cell division. Over-expression of sigE reduced respiration and photosynthesis activities via changes in gene expression and chlorophyll fluorescence. Hydrogen production under micro-oxic conditions is enhanced in sigE over-expressing cells. Despite these pleiotropic phenotypes, the sigE over-expressing strain showed normal cell viability under both nitrogen-replete and nitrogen-depleted conditions. These results provide insights into the inter-relationship among metabolism, cell morphology, photosynthesis and hydrogen production in this unicellular cyanobacterium.

Introduction

The requirement for renewable energy has increased attention on metabolic engineering of photosynthetic organisms including cyanobacteria. Many researchers have attempted to modify metabolic enzymes for production of metabolites of interest. However, metabolic reactions may not be the bottleneck for metabolite production. Cell morphology and photosynthetic electron transport should also be considered, because primary metabolism is tightly associated with photosynthesis (Ducat et al., 2011).

Synechocystis sp. PCC 6803 (hereafter Synechocystis 6803) is a non-nitrogen-fixing unicellular cyanobacterium, and is the most widely investigated cyanobacterium species. Cyanobacteria perform oxygenic photosynthesis and are the only prokaryotes known to possess circadian rhythms (Kondo et al., 1993). Synechocystis 6803 cells are spherical and approximately 1.5 μm in diameter (van de Meene et al., 2006). The ultrastructure of Synechocystis 6803 was resolved by means of three-dimensional imaging (van de Meene et al., 2006), and the biogenesis and dynamics of multiple membrane structures have been intensively explored (Nickelsen et al., 2011). Genetic and biochemical aspects of cell division in Synechocystis 6803 have been investigated (Miyagishima et al., 2005; Marbouty et al., 2009), but the involvement of transcriptional regulators in the regulation of cell division remains unclear.

Photosynthetic activity may be measured using a Clark-type oxygen electrode or by chlorophyll fluorescence analysis (Campbell et al., 1998). The latter method enables quantification of the maximal photochemical efficiency of photosystem II (PSII) (Fv/Fm), the photochemical efficiency of open PSII centers under a given light acclimation status (inline image), photochemical quenching (qP), non-photochemical quenching (qN and NPQ), and the effective quantum yield of electron transport through PSII (ΦII) as empirically verifiable indices of photosynthetic status (Campbell et al., 1998). During nitrogen starvation, phycobilisomes (light-harvesting antennae comprising rod and core proteins) are degraded, and photosynthetic activities continuously decrease (Schwarz and Forchhammer, 2005). Overall, the genes encoding photosynthetic electron transport are down-regulated under nitrogen starvation (Osanai et al., 2006; Aguirre von Wobeser et al., 2011; Krasikov et al., 2012).

NtcA, a transcription factor that is conserved among all cyanobacterial species, controls gene expression related to nitrogen and carbon assimilation during nitrogen starvation (Vega-Palas et al., 1992; Herrero et al., 2001). NtcA also regulates the expression of several transcriptional regulators, including SigE, an RNA polymerase sigma factor whose expression is induced by nitrogen depletion (Muro-Pastor et al., 2001). Genes involved in sugar catabolic pathways, such as glycogen catabolism, glycolysis and the oxidative pentose phosphate pathway, are down-regulated by knockout of sigE (Osanai et al., 2005a,b, 2008). Disruption of sigE expression reduces Synechocystis 6803 cell viability under various conditions such as prolonged nitrogen starvation, light-activated heterotrophic conditions, mixotrophic growth conditions and salt stress (Muro-Pastor et al., 2001; Osanai et al., 2005b, 2009; Summerfield and Sherman, 2007; Pollari et al., 2008). Previously, we showed that over-expression of sigE enhances sugar catabolism under normal growth conditions and polyhydroxybutyrate production under nitrogen starvation (Osanai et al., 2011, 2013). Microarray analysis showed that SigE also up-regulates expression of genes encoding hydrogenases (Osanai et al., 2011).

In this study, we demonstrate the pleiotropic effect of sigE over-expression in Synechocystis 6803 on cell morphology, photosynthesis, hydrogen production and the protein levels of various regulators. Despite these changes, the sigE over-expressing strain shows similar viability to that of the parental glucose-tolerant (GT) wild-type strain, which presents an opportunity for genetic engineering of Synechocystis 6803 without affecting cell proliferation.

Results

Comparative cell size of the sigE over-expressing and parental wild-type strains

Transmission electron microscopy revealed that the cells of the sigE over-expressing strain were larger than those of the GT strain (Figure 1a). The cell diameters of the GT and sigE over-expressing strains were 1.57 ± 0.22 and 2.52 ± 0.42 μm, respectively (mean ± SD). Cell division in the GT strain generated two daughter cells of similar size (Figure 1b). In contrast, division of sigE over-expressing cells sometimes showed unequal cleavage (Figure 1c). Most thylakoid membranes in GT cells were located at the periphery, close to the cytoplasmic membrane (Figure 1b). However, some thylakoid membranes in the sigE over-expressing strain traversed the central cytoplasm (Figure 1c). Scanning probe microscopy also showed the increased cell size of the sigE over-expressing strain (Figure 2). A structure that resembled a cell division ring was frequently observed in the sigE over-expressing strain, but this feature was unclear in GT cells (Figure 2). The ring-like structure remained ambiguous in GT cells under increased magnification (Figure 2).

Figure 1.

Transmission electron micrographs of GT and sigE over-expressing (GOX50) cells grown in nitrogen-replete medium. (a) Micrographs of GT and sigE over-expressing cells at the same magnification. Scale bars = 5 μm. (b,c) Micrographs of GT cells (b) (scale bar = 0.5 μm) or sigE over-expressing cells (c) (scale bar = 1 μm) undergoing cell division. Black dots indicate glycogen granules. White and gray arrowheads indicate polyphosphate bodies and polyhydroxybutyrate granules, respectively. Black arrows indicate the invaginating septum. sigE over-expression caused aberrant cell morphologies, in which the second cell division started before the first cell division was completed.

Figure 2.

Scanning probe micrographs of GT and sigE over-expressing (GOX50) cells grown in nitrogen-replete medium. (a) GT cells and (b) sigE over-expressing cells that were dried on a cover glass. The right panels show higher-magnification images. Black arrows indicate cell division rings.

sigE over-expression decreases photosynthetic electron transport

Respiration was decreased in the sigE over-expressing strain grown under both nitrogen-replete and nitrogen-depleted conditions for 1 day (Table 1). Photosynthetic activity in the sigE over-expressing strain was lower than that in the GT strain under nitrogen-replete conditions, but was higher than that in the GT strain after 1 day of nitrogen depletion (Table 1). Measurement of photosynthetic activity under various light intensities revealed that oxygen evolution in the sigE over-expressing strain was lower than that in the GT strain only under high-light conditions (Figure 3).

Figure 3.

Photosynthetic activities of GT and sigE over-expressing (GOX50) cells under various light intensities. Values are means ± SD from six independent experiments. Differences between GT and sigE over-expressing cells were analyzed by Student's t test. Asterisks indicate statistically significant differences compared the GT with GOX50 strain at *< 0.05 and **< 0.005.

Table 1. Respiration and total photosynthetic activities in the GT and sigE over-expressing (GOX50) strains under nitrogen-replete (+N) and nitrogen-depleted (−N) conditions
ParameterRespiration (μmol O2 mg−1 Chl a h−1)Photosynthetic activities (μmol O2 mg−1 Chl a h−1)
  1. Photosynthetic activities were measured under white light illumination (1050 μmol photons m−2 sec−1). Values are means ± SD from seven independent experiments. Differences between GT and sigE over-expressing strains were analyzed by Student's t test. Asterisks indicate statistically significant differences compared the GT with GOX50 strain at *< 0.05 and **< 0.005.

GT
+N17.2 ± 5.653.1 ± 3.9
−N (1 day)27.9 ± 2.430.9 ± 6.9
−N (3 days)18.9 ± 1.67.7 ± 2.3
GOX50
+N14.0 ± 2.6*45.8 ± 4.3*
−N (1 day)22.4 ± 2.1**39.4 ± 6.8**
−N (3 days)18.6 ± 2.48.3 ± 1.1

Chlorophyll fluorescence analysis revealed that ΦII and inline image were lower in the sigE over-expressing strain under nitrogen-replete conditions, but were similar between the two strains under nitrogen depletion (Figure 4). Fv/Fm was slightly lower in the sigE over-expressing strain only under nitrogen starvation (Figure 4). Both qN and NPQ were higher in the sigE over-expressing strain than in the GT strain under nitrogen-replete conditions (Figure 4). Under nitrogen starvation, neither qN nor NPQ decreased in the GT strain, whereas both parameters were reduced in the sigE over-expressing strain (Figure 4).

Figure 4.

Photosynthetic parameters for GT and sigE over-expressing (GOX50) cells derived from chlorophyll fluorescence analysis. Values are means ± SD from six independent experiments. Differences between GT and sigE over-expressing cells were analyzed by Student's t test. Asterisks indicate statistically significant differences compared the GT with GOX50 strain at *< 0.05 and **< 0.005.

The expression of genes that encode cytochrome c oxidases (CI, DI, DII, EI and EII) and cytochrome c oxidase folding protein (ctaB) was induced under nitrogen deprivation and down-regulated by sigE over-expression (Figure S1). All eight genes that encode proteins located at the center of the PSII dimer (psbA2/A3, B, C, D1, D2, E, F and I) were down-regulated by sigE over-expression (Figure S2). In terms of PSII peripheral proteins, only three of the nine genes (psbJ, L and Z) were repressed by sigE over-expression under nitrogen-replete conditions (Figure S3). Only psbO transcription was reduced in the sigE over-expressing strain among genes encoding components of the oxygen-evolving complex of PSII (PsbO, U and V) (Figure S4). Eight of 12 genes encoding photosystem I (PSI) subunits (psaC, E, F, I, J, K1, K2, L and M) were down-regulated by sigE over-expression under either nitrogen-replete or nitrogen-depleted conditions (Figure S5). In contrast to the transcript levels, the PsbA and PsbO protein levels were similar in both strains, whereas the PsaD protein levels were increased by sigE over-expression (Figure S6).

sigE over-expression increases hydrogen production under micro-oxic conditions

Quantitative real-time PCR demonstrated that transcripts of five genes encoding hydrogenases (hoxEFHUY) increased by 1.7–3.1 times upon sigE over-expression (Figure 5). It is known that hydrogen accumulates under micro-oxic conditions (Schütz et al., 2004). The cells were subjected to micro-oxic conditions under N2 gas for 1 h, and sealed in a gas chromatography (GC) vial for 1 day under light or dark conditions with or without external carbon sources (Na2CO3 or NaHCO3) (Figure 6). Overall, the hydrogen concentration was higher under dark conditions than that under light conditions (Figure 6). The amount of hydrogen produced by the sigE over-expressing strain was approximately twice that produced by GT strain under both light and dark conditions (Figure 6). Addition of Na2CO3 or NaHCO3 inhibited hydrogen production under both conditions (Figure 6). Hydrogen accumulation was diminished in the GT strain under light conditions in the presence of NaHCO3, but hydrogen was detected in the sigE over-expressing strain under same conditions (Figure 6). The maximum concentration of accumulated hydrogen in this study was 1.5% (1526 ppm), which was the value for the sigE over-expressing strain incubated under dark conditions without carbon sources (Figure 6).

Figure 5.

Transcript levels of genes encoding hydrogenases in GT and sigE over-expressing (GOX50) cells. Values are means ± SD from six independent experiments. The levels were calibrated relative to the value obtained in the GT strain under nitrogen-replete conditions, which was set at 100%. Differences between GT and sigE over-expressing cells were analyzed by Student's t test. Asterisks indicate statistically significant differences compared the GT with GOX50 strain at *< 0.05.

Figure 6.

Hydrogen accumulation during micro-oxic conditions in GT and sigE over-expressing (GOX50) cells. The upper part shows the cell density and air and light conditions during incubation. Cells were suspended in 10 ml medium. The graphs show the hydrogen concentration in a 10 ml headspace in a GC vial under light (left) or dark (right) conditions in the absence or presence of 100 mm Na2CO3 or 100 mm NaHCO3. The maximum scale values for the y axis are 45 ppm (light conditions) or 2200 ppm (dark conditions). Values are means ± SD from three or four independent experiments. Differences between GT and sigE over-expressing cells were analyzed by Student's t test. Asterisks indicate statistically significant differences compared the GT with GOX50 strain at *< 0.05.

Interference of regulatory protein levels in the sigE over-expressing strain and measurement of viability

The effect of sigE over-expression on the levels of regulatory proteins was examined (Figures S7 and S8; see detailed information in Appendix S1). Three proteins, Rre37, Rre12 and PII, which were induced by nitrogen depletion, showed reduced accumulation in the sigE over-expressing strain (Figure S7b). The level of the SigE-interacting protein ChlH decreased in the sigE over-expressing strain after 6 h of nitrogen depletion (Figure S8). In contrast, the protein level of SigA, a primary sigma factor, was increased by sigE over-expression irrespective of nitrogen status (Figure S8).

Cell viability was measured by spotting cells previously grown under nitrogen-replete or nitrogen-depleted conditions onto modified BG-11 medium. Irrespective of the pleiotropic phenotypes, cell viability under both nitrogen-replete conditions and nitrogen-deficient conditions (3 or 7 days) was similar in the GT and sigE over-expressing strains (Figure 7).

Figure 7.

Viability of GT and sigE over-expressing (GOX50) cells under nitrogen-replete and nitrogen-depleted conditions. Cultures of GT or sigE over-expressing cells grown to the mid-exponential phase under nitrogen-replete conditions (left) or nitrogen-depleted conditions for 3 days (middle) or 1 week (right) were spotted onto BG-11 agar plates containing 10 mm NH4Cl as a nitrogen source. Each spot consisted of 2 μl of culture diluted to an A730 of 1.0, 0.5, 0.1 or 0.05, as indicated. The plates were photographed after incubation for 6 days under light conditions (approximately 50–70 μmol photons m−2 sec−1).

Discussion

In this study, we have shown that sigE over-expression leads to various phenotypes in Synechocystis cells (Figure 8), and these results suggested a close relationship among metabolism, photosynthesis, cell morphology and hydrogen production. Among marine phytoplankton, small species, which show faster growth, contain a decreased number of PSII centers compared with large species (Key et al., 2010). Cell division is directly regulated by metabolism in some bacteria. Bacillus subtilis cells carrying a loss-of-function mutation in pgcA, which encodes a phosphoglucomutase, were approximately 35% shorter than wild-type cells (Weart et al., 2008). Biochemical analysis revealed that UgtP, an UDP-glucose transferase, inhibits FtsZ assembly in vitro (Weart et al., 2008). In Escherichia coli, the knockout mutant of phosphoglucomutase has shorter cells (Lu and Kleckner, 1994). Our analyses reveal that levels of UDP-glucose, UDP-galactose, UDP-glucuronate and GDP-mannose are increased by sigE over-expression under nitrogen-replete conditions (Osanai et al., 2013). Aberrant cell division in the sigE over-expressing strain may be caused by these changes in the levels of sugar nucleotides.

Figure 8.

Summary of the pleiotropic effect of sigE over-expression in Synechocystis sp. PCC 6803. The major effects of sigE over-expression are categorized under cell morphology, photosynthesis, and hydrogen production.

The cell size of Synechocystis 6803 is increased in mutants that lack aqpZ, hik31 or abrB (Akai et al., 2011; Yamauchi et al., 2011; Nagarajan et al., 2012). All three mutants show decreased expression of genes associated with either sugar catabolism or carbon assimilation (Lieman-Hurwitz et al., 2009; Akai et al., 2011; Nagarajan et al., 2012), which suggests that cell size and sugar metabolism are correlated in Synechocystis 6803. In terms of cell division-related proteins, the transcript level of ftsZ, which encodes a protein that forms a Z-ring during cell division, is decreased by knockout of abrB or sigE over-expression (Ishii and Hihara, 2008; Osanai et al., 2011). The expression of other components associated with cell division as described by Marbouty et al. (2009) and Miyagishima et al. (2005) was unaffected by sigE over-expression (Osanai et al., 2011). Ring-like structures were clearly observed in the sigE over-expressing strain (Figure 2), which implied that cell division was retarded by sigE over-expression, although we cannot exclude the possibility that the ring-like structures in the sigE over-expressing cells were discernible because of the increased cell size. Although the cause of the increased cell size remains unclear, these results indicate that cell division is coordinately regulated by the sugar metabolism status in this cyanobacterium.

Over-expression of sigE resulted in reduced oxygen evolution under nitrogen-replete conditions (Table 1). Empirically, ΦII and inline image are positively correlated with oxygen evolution (Campbell et al., 1998), consistent with our results showing that both ΦII and inline image decreased in the sigE over-expressing strain under nitrogen-replete conditions (Figure 4). PSII genes were down-regulated by sigE over-expression (Figures S1–S5), and thus oxygen evolution was decreased by sigE over-expression at the transcriptional level under nitrogen-replete conditions. Given that qN and NPQ increased (Figure 4), sigE over-expression may result in over-reduction in the cells. SigE activates the enzymatic activities of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase (Osanai et al., 2011), which reduce NADP to NADPH, leading to an increased NADPH/NADP ratio in the sigE over-expressing strain. The finding that only the sigE over-expression strain produced hydrogen under light conditions in the presence of NaHCO3 (Figure 6) may also indicate the presence of increased reductants in this strain.

After 1 day of nitrogen deprivation, oxygen evolution by the GT and sigE over-expressing strains decreased to 58 and 86%, respectively, compared with nitrogen-replete conditions (Table 1). The levels of oxygen evolution were similar between the strains after 3 days of nitrogen depletion (Table 1), indicating delayed down-regulation of photosynthesis in the sigE over-expressing strain during nitrogen starvation. The respiratory activity of both two strains increased 1.6-fold after 1 day of nitrogen depletion (Table 1). The transcript levels of PSII and PSI genes were similarly repressed in the two strains under nitrogen starvation (Figures S2–S5), suggesting that the photosystems of the GT and sigE over-expressing strains were only different at the post-transcriptional level after nitrogen starvation. Both qN and NPQ were reduced by sigE over-expression after 1 day of nitrogen depletion (Figure 4). The reduced non-photochemical quenching may lead to increased linear photosynthetic electron transport because excess energy cannot be thermally dissipated. Using triple mutant lines, it was shown that the strain containing an increased level of SigE proteins relative to other sigma factors exhibited sensitivity to photoinhibition (Pollari et al., 2011), consistent with the idea that sigE over-expression led to aberrant thermal dissipation. Detailed analysis is required to reveal the biochemical properties of the photosystems of the two strains after nitrogen depletion.

In an optimal hydrogen production system, all electrons derived from water splitting in PSII are directed to hydrogenase (Tamagnini et al., 2007). Hydrogen production is thus competitive with respiration and the Calvin cycle, and hydrogenase is sensitive to oxygen (Tamagnini et al., 2007). Our results are consistent with this idea, as hydrogen production was decreased by illumination or addition of carbonate/bicarbonate (Figure 6). As hydrogen production is tightly coupled to photosynthesis and metabolism, it is plausible that the expression of genes that encode sugar catabolic enzymes, photosynthetic proteins and hydrogenases is concomitantly regulated by a transcriptional regulator (in this case SigE). Two AbrB-like transcription factors in Synechocystis 6803 regulate expression of both the hydrogenase operon (Oliveira and Lindblad, 2008; Dutheil et al., 2012) and enzymes involved in carbon and nitrogen metabolism (Ishii and Hihara, 2008; Lieman-Hurwitz et al., 2009), also suggesting a close relationship between hydrogen production and primary metabolism.

The levels of several regulatory proteins were altered by sigE over-expression (Figures S7 and S8). At present, however, we are unable to link the changes in protein levels and the pleiotropic phenotypes of sigE over-expressing cells. Some phenotypes observed by sigE over-expression may be an indirect effect because several transcriptional regulators were altered in this strain (Figures S7 and S8). The decreased levels of nitrogen regulators PII and Rre37 (Figure S7) are one possible reason for the delayed response to nitrogen starvation, including the higher photosynthetic activity after 1 day of nitrogen starvation by sigE over-expression (Table 1). The levels of these regulatory proteins appear to be mutually regulated (discussed in Appendix S1). Despite these changes in sugar metabolism, cell morphology, photosynthesis and hydrogen production, it is surprising that cell viability under nitrogen-replete and nitrogen-depleted conditions was similar to that of the parental wild-type cells (Figure 7). Although the effect of sigE over-expression on some phenotypes may be indirect, genetic analysis of the sigE mutant will be beneficial to obtain an improved understanding of the inter-relationships between metabolism, cell morphology, photosynthesis and hydrogen production in this cyanobacterium, and to perform genetic and metabolic engineering without affecting cell proliferation in the future.

Experimental Procedures

Bacterial strains and culture conditions

The glucose-tolerant (GT) strain of Synechocystis sp. PCC 6803, isolated by Williams (1988), and the sigE over-expressing strain, named GOX50 (Osanai et al., 2011), were grown in BG-110 liquid medium (Rippka, 1988) containing 5 mm NH4Cl (buffered with 20 mm HEPES/KOH, pH 7.8), termed modified BG-11 medium. Among the GT sub-strains, the GT-I strain was used in this study (Kanesaki et al., 2012). Liquid cultures were bubbled with 1% v/v CO2 in air, and incubated at 30°C under continuous white light (approximately 50–70 μmol photons m−2 sec−1). For plate cultures, modified BG-11 medium (containing 10 mm NH4Cl in liquid medium) was solidified using 1.5% w/v agar, and the cultures were incubated in air at 30°C under continuous white light (approximately 50–70 μmol photons m−2 sec−1). For this strain, the modified BG-11 liquid medium was supplemented with 10 μg ml−1 kanamycin (Sigma-Aldrich, http://www.sigmaaldrich.com/united-states.html) during pre-culture. Growth and cell densities were measured by determining the absorbance at 730 nm (A730) using a Hitachi U-3310 spectrophotometer (Hitachi, http://www.hitachi.com/). For nitrogen-starved conditions, cells were collected by filtration through a mixed-cellulose ester membrane (Advantec, http://www.advantec.co.jp/en/) and resuspended in BG-110 liquid medium.

Electron microscopy

Cells from mid-exponential-phase cultures of Synechocystis 6803 (A730 = 0.5–0.8) grown in modified BG-11 medium were collected by centrifugation at 1000 g for 10 min. The cells were fixed for transmission electron microscopy as described by Akai et al. (2011). The ultra-thin sections were examined using a transmission electron microscope (JEM-1400; JEOL Ltd., http://www.jeol.co.jp/en) at 80 kV.

Scanning probe microscopy

Cells grown on BG-11 agar plates were suspended in 1 ml sterilized water and centrifuged at 20 500 g for 1 min. The supernatant was discarded, and the cells were washed twice with 1 ml sterilized water. The cells were resuspended in 1 ml sterilized water, and 10 μl of the cell suspension was spotted onto a cover glass and dried at 96°C for 1 min. The cells were examined using a scanning probe microscope (SPM-9700; Shimadzu, http://www.shimadzu.com/) according to the manufacturer's instructions.

Measurement of respiratory and photosynthetic activities

Chlorophyll contents of cells grown under nitrogen-replete conditions were measured using a methanol extraction method (Grimme and Boardman, 1972). Cells containing 10 μg chlorophyll were resuspended in 1 ml BG-110 liquid medium, supplemented with or without 5 mm NH4Cl, and incubated at 30°C within the chamber of an Oxytherm Clark-type oxygen electrode (Hansatech Instruments, http://hansatech-instruments.com/). Cells were adapted to dark conditions with monitoring of oxygen consumption for 10 min. The rate of oxygen consumption in the final 3 min of incubation was used to determine respiration activity. Total oxygen evolution was measured after addition of 10 μl of 1 m NaHCO3 and exposure to white light of various intensities. The rate of oxygen evolution was calculated for the final 3 min of the 7 min measurement period.

Chlorophyll fluorescence

Chlorophyll fluorescence was measured using an AquaPen-C AP-C 100 fluorometer (Photon Systems Instruments, http://psi.cz/). Chlorophyll contents of cells grown under nitrogen-replete and nitrogen-depleted conditions were measured, and cells were diluted to 0.3 μg ml−1 chlorophyll a in 2 ml BG-110 medium supplemented with or without 5 mm NH4Cl. After dark incubation for 5 min, the chlorophyll fluorescence was measured according to the manufacturer's instructions provided by Photon Systems Instruments (protocol NPQ1). The intensities of actinic light and pulse-saturated light were 300 and 1500 μmol photons m−2 sec−1, respectively. The wavelength of actinic light and pulse-saturated light was 450 nm. The Fm value was obtained after addition of 10 μm 3-(3,4-dichlorophenyl)-1,1-dimethylurea) (DCMU). The values of photosynthetic parameters were calculated as described previously (Campbell et al., 1998; Sonoike et al., 2001), except that far-red light was not used in the present experiment. The values for qP, qN, NPQ and ΦII were calculated as inline image, respectively.

RNA isolation and quantitative real-time PCR

RNA isolation was performed as described previously (Osanai et al., 2013). The cDNAs were synthesized using the SuperScript III first-strand synthesis system (Life Technologies, http://www.lifetechnologies.com) with 2 μg total RNA. Quantitative real-time PCR was performed using the StepOnePlus real-time PCR system (Life Technologies) according to the manufacturer's instructions, using the primers listed in Table S1. The expression level of rnpB, which encodes RNaseP subunit B, was used as an internal standard.

Gas chromatography/thermal conductivity detector analysis

Pre-cultured cells were diluted to A730 = 0.4 in 70 ml modified BG-11 medium. Cells were cultured for 1 day, and collected by centrifugation at 10 000 g for 2 min, followed by concentration to A730 = 2.0 in 10 ml BG-11 medium modified using 500 mm HEPES/KOH (to avoid pH changes resulting from use of NaCO3) in a 20 ml GC vial (leaving 10 ml headspace). The GC vial was filled with N2 gas introduced using a GC syringe for 1 h. After removal of the syringe, the sealed vial was incubated at 30°C under continuous white light (approximately 50–70 μmol photons m−2 sec−1) or in the dark for 1 day. The accumulated H2 gas was measured using a gas chromatograph (GC-2010 Plus AT; Shimadzu) according to the manufacturer's instructions. N2 was used as the carrier gas with a flow rate of 10 ml min−1.

Statistical analysis

All statistical analyses were performed using StatPlus:mac LE software for MacOSX (AnalystSoft, http://www.analystsoft.com/en/). Paired two-tailed t tests were performed to calculate P values with a 95% significance level.

Acknowledegments

This work was supported by the Ministry of Education, Culture, Sports, Science and Technology, Japan, a Grant-in-Aid to T.O. for Precursory Research for Embryonic Science and Technology (project name ‘Production of cyanobacterial bioplastics by metabolic engineering with modified carbon dynamics’) from the Japan Science and Technology Agency, and by a Grant-in-Aid to M.Y.H. for Creative Research for Embryonic Science and Technology (project name ‘Elucidation of amino acid metabolism in plants based on integrated omics analyses’). The authors also wish to acknowledge the editorial assistance provided by the Edanz Group Japan in preparation of this manuscript (language-editing service).

Ancillary